Natural-language querying of parliament’s website
Public engagement and open parliament
Italy - Senate
Use case ID: 024
Author: Senate of Italy
Date: 12 June 2024
Objective:
Enable users to query the Senate website’s search engines using natural language processed by large language model (LLM)-based artificial intelligence (AI), enhancing the accessibility, accuracy and user experience of the search functionality.
Actors:
- Senate website users (citizens, researchers and journalists)
- Senate IT and web development team
Prerequisites:
- Existing search engine integrated with the Senate website
- Trained LLM-based AI model for understanding and processing natural-language queries
- Database of Senate documents and information to be searched
- Internet accessibility for users
Scenario:
- The user accesses the Senate website.
- The user enters a query in natural language (e.g. “What are the latest laws passed on education?”) in the search bar.
- The LLM-based AI model processes the natural-language query to understand the intent and key terms.
- The LLM-based AI model translates the natural-language query into a format that the search engine can understand.
- The search engine retrieves relevant documents and information based on the translated query.
- The results are displayed to the user in a user-friendly format.
- The user can refine their search or ask follow-up questions in natural language to further narrow down the results.
Alternate flows:
- If the LLM-based AI model cannot understand the query, it prompts the user to rephrase or provides suggestions.
- If the search yields too many or too few results, the system offers advanced search options or filters to refine the search.
Expected results:
- User satisfaction is improved owing to more accurate and relevant search results.
- Usage of the Senate website’s search functionality is increased.
- It takes less time and effort for users to find specific information.
- Accessibility for users unfamiliar with technical search terms or Senate document classifications is improved.
Potential challenges:
- None
Data requirements:
- Historical search queries and user interactions for training and improving the LLM model
- Database of Senate documents, laws and other relevant information
Integrations with other systems:
- Existing search engine infrastructure of the Senate website
- LLM-based AI processing systems and models
- User interface components for displaying search results
Success metrics:
- Query response time
- User satisfaction ratings and feedback
- Accuracy and relevance of search results
- Increase in the number of natural-language queries
- Reduction in user queries requiring manual intervention or support